Assessment of Offshore Wind Resources, Based on Improved Particle Swarm Optimization

Abstract

It is crucial to understand the characteristics of wind resources and optimize wind resources in the area that is being considered for offshore wind farm development. Based on the improved particle swarm optimization (IPSO) and the back propagation neural network (BPNN), the IPSO-BP hybrid intelligent algorithm model was established. The assessment of wind resource characteristics in the eastern waters of China, including average wind speed, extreme wind speed, wind power density, effective wind energy hours and wind direction distribution were all calculated. Additionally, the wind speed throughout the different years in Luchao Port, a famous seaport in China, was predicted. The results revealed that the wind power density is approximately 300 W/m2 all year round and that the effective wind energy hours take up about 92% per hour. It was also identified that the wind direction distribution is stable in Luchao Port, implying that there are better wind energy resource reserves in this region. The IPSO-BP model has a strong tracking performance for wind speed changes, and can accurately predict the wind speed change in a short period. In addition, the prediction error of the IPSO-BP model is smaller when the time of training data is closer to the target one, and it can be controlled within a 5% range.